Registered users: Bing [Bot], Majestic-12 [Bot], Yahoo [Bot]
Who is online
In total there are 78 users online :: 3 registered, 0 hidden and 75 guests
Most users ever online was 218 on Wed Dec 07, 2016 6:58 pm
Registered users: Bing [Bot], Majestic-12 [Bot], Yahoo [Bot] based on users active over the past 5 minutes
As you may or may not be able tell from the title I plan to build a railgun (or Linear Electromagnetic Accelerator) for a science fair project. I was recently accepted into the South West Virginia's Governors School. They hold a Science fair there every year entering a railgun as my project.
My idea came the railgun built at Power Labs and http://www.geocities.com/physicsthings/microrailgu.htm
I planned to construct the railgun with a few idea in mind:
1. I want to see what effect(affect?) super conductor projectiles have on the performance of the gun(i.e. rail erosion, projectile speeds, arcing, and plasma formation) I just to dunk the projectile in liquid N2 prior to loading.
2. I wanted to experiment with active cooling the rails with either non-electrically conductive heatsinks, dipping the rails in liquid N2, or some sort of liquid nitrogen cooling apparatus
3. I want to use a hybrid launcher as the injector to achieve the highest possible injection velocity that I can.
I want to build the injector from an 8" piece of sch80 steel pipe and run at a 10x mix. The "barrel" will be a 12" long square tube the will 1/4" by 1/4" with two sides made of Teflon to cut down on friction. The included diagram should explain. I need to know the burst pressure of 3" Sch80 steel pipe.
Also, would a refrigerator compressor be able to pump liquid N2, or would it freeze it up?
I plan on building the railgun atleast 4' long with 1' as the injector barrel with bollts place every two inches as per the diagram. I've made some design speculations from the information from the PowerLabs website and the design of the railgun will most likely change before this is all said and done.
I'm not exactly sure what kind of capacitors I'm going to need but I know I need at least 30 400V caps maybe more. That's all for now.
I will also try and add a picture of the HGDT simulation I created.
Last edited by strangers on Fri May 08, 2009 6:59 pm, edited 1 time in total.
For the high current you are talking, you will want low inductance types. Look for a capacitor listed as a "Pulse" capacitor. They are built internally to handle the high magnetic forces from the very high currents that would rip other capacitors apart. If you search the web for "Magnetic Can Crusher" or "Coin Shrinking" you may find some info on these very high current devices.
Here is an example page;
In regards to the capacitor used;
Warning, they are not cheap.
sch 80 3" has a working pressure of 768 psi.
my calculated burst pressure is 10,000psi
10x hybrid usually provides max 1,000psi.
Do you want to bump it to 20-35x with that fridge compressor? 35x means 3500psi max. A longer barrel would dramatically improve injector performance.
The muzzle blast from the injector could possibly warm the projectile and or rails above superconducting temperature. I can't think of any strong, non conductive materials that don't get extremely brittle at LN temperatures. I vote for just dropping superconducting, since a majority of resistance comes from the ESR of the capacitors and poor contact between the projectile and rails.
If you still want continuously pumped LN, have the draw from the LN below the surface of the liquid and let the boiling of the LN produce pressure and just force it into said tube (gas tight flask would be required).
EDIT: here's some pulse caps if you are interested.
In case you live in Dallas
I think what I'll do is have the entire railgun be 7' long, 4' for the railgun and 3' for the injector. As per the HGDT sim, this should be able to accelerate it enough so that the railgun can bring it past the speed of sound. As for the cooling what I want to do is have the fridge compressor just pump air through a metal(probably copper) line that would have one coil in the liquid nitrogen(maybe even just dry ice) or above it, and one coil wrapped around the railgun and the hybrid chamber to avoid overheating the rails and burning myself. Here's an updated diagram.
don't bite off more than you can chew.... that is MASSIVE. What energy do you plan on using? If you go through with this, it will cost ALOT. I made a railgun enclosure out of G-11, 2' long and a 2"x1.5" cross section, and it cost 370$ (custom machined).
I do not think you are able to build this size.
The fact you wan't to use a injector. Big railgun's like this will be killed by a injector, a large SCR bank is byfar a better choice.
superconducting is a bad choice. You cannot just dump some copper in LN2 and expect it to be superconductive. You need special alloy's (costs ALOT). Also, just the friction and sparking alone will almost instantly bring the projectile to non-superconducting.
400v caps says nothing. I used 18 400v 3900uf capacitors for 5.6kJ. Even if you went with 20kj, the rail length only needs to be 1 foot long. The pulse time is SHORT, you need to run some proper LCR sims.
Past the speed of sound is easy. Mine comes very close with that power level, that size is like MACH 5, if you design it correctly.
EDIT: There is also the problem of material strength at that low temperatures. Powerlabs is misleading in lots of aspects, mainly the Inductor. He states that you must have a injector, but that is FALSE. Think about it. A railgun is practically a short. As soon as the projectile touches the rails, current flows. Most of the power will be gone before the projectile gets in the rails.
My railgun is at:
I would actually recommend an inductor, to keep the first few rails from getting destroyed. IIRC, the railgun is more efficient further down the rails (away from the contact points), so get some really heavy wire and wind it around a non conductive core.
For more efficiency, you could run cables parallel to the rails along the hybrid barrel section (see pic) make sure to secure them VERY WELL. for variables, you could change injection velocity, or add external magnets to the rails.
Also, look into those 'lytics in my last post; I use 12 for a ring launcher (4seriesx3parallel). Needless to say, I want MOAR.
EDIT: uploaded the pic I talked about
I disagree with the inductor. The goal is to increase the action integral, which can be done by highering power, or lowering resistance and inductance, or both. Frankly, lowering the current that little is not gonna do wonders on efficiency, I have tried it with mine. Running the cable parallel backwards will help, but not much. Keep in mind that the entry part of the rails needs to be insulated, with copper underneath. This ensures that the magnetic field starts behind the rails.
BTW, I have a extra G-11 railgun enclosure if your interested. Exactly the same as the one in my site, 1.5" x 2" cross section, two feet long, 1 foot injection. Held together by around 24 1/4" bolts, machined to .001" (2 piece set)
I think you guys are right I just went back through all the math and figured that I might go with a 3' or 2' gun for the cost.
But to further reduce the cost I might use electrical grade fiberglass, but anything that needs to be cut I can do at my school's shop, so no machining costs. A 6'x1/2"x3" piece of electrical grade fiberglass for Mcmaster-Carr is only $26. And it's red, that's a plus.
@Ramses: Do those electrolytic caps from BCD discharge as fast as say, pulse caps or will it actually make a big difference?
@rp181: What gauge wire did you use to connect you cap bank to the rails, I was planning on 10awg but I could go lower.
What do you guys recomend to charge a cap bank of say 24 or 36 450V 2400uF capacitors, or atleast what kind of chargers do you use?
I hope you guys don't mind pickin' your brains.
Excellent question. At high pulse currents such as used in rail guns, can crushers, and coin shrinkers, the very high currents must be accounted for. The mechanical forces from the magnetic fields is extreme. Caps ripping them selves apart is only part of the problem. All the conductors, including those inside the can must be able to handle the high current. How the conductor is attached to the plates is critical so the attachment point does not behave like a foil fuse and instantly vaporise. The sudden creation of very high temperature metal vapor can rip the cap apart.
These discharge energies are best delivered to the rail, not inside the capacitor bank.
Viewing the terminals on the capacitors in the link show the terminals are NOT designed for high energy pulse applications. These are most likely high voltage power supply capacitors as used in broadcast transmitters. (worked on a lighting damaged FM radio station that used these.)
High voltage high current capacitors look more like these.
Note the size and location of the terminals as well as the terminal support.
These are transmitter capacitors.
The caps ripping apart are not a problem, the energy is added. You do need to worry about the main bars though. If you look under my build log, you can see where the two parrallel bars stick out. During a shot, those jarred apart, causing massive losses from poot contact. Make sure you use a corrosion inhibitor.
I used 00 gauge wire, 10 guage is way too small. I suggest storm copper. All my connections are 1"x1/8" aluminium bar.Fiberglass was my first choice, but i had the oppurtunity to get these. I suggest the 6"x.5"x3' fiberglass piece, its about the same price.
High current capacitors are different from pulse capacitors. High current are designed for constant current, at up to 600A. Those types are used in high frequency pulse, and induction heating. Lyctics are fine for a hobbyiest.
Capacitors are expensive. I was lucky to get each of mine for 7$, that sale was stopped recently.
you can keep pickin my brain =)
I figured I might as well post my bill of materials and see if anyone can see anything I might be missing(aside from materials for the charger and the flange, and stun gun, which I already have):
x1 Fiberglass sheet - http://www.mcmaster.com/#3345k44/=1ug10h
x1 Fiberglass sheet - http://www.mcmaster.com/#3345k26/=1ugmbc
x2 Copper Strips - http://www.mcmaster.com/#8964k291/=1ugdgs
x2 Teflon strips - http://www.mcmaster.com/#8545k154/=1ugdpr
x30 18-8 stainless steel bolts -
x60 washers -
x30 nuts -
x30 Capacitors - http://stores.shop.ebay.com/BCD-Parts
00ga Wire -
x3 Aluminum buss bars - http://www.mcmaster.com/#8975k527/=1ugqnk
8” of 3” sch80 steel pipe
x1 spark plug
x1 spark plug cap
x1 200 psi pressure gauge
x2 1000 psi valves
x1 sch120 1" steel union
x1 rerigerator compressor
oh my god. Why so much material? The first 4 items is more than twice what i spent. With that kind of materials, you should just buy my enclosure =)
Everything does look good though. If you have the money for that kind of stuff, ditch the injector, buy a SCR assembly.
How come everyone can find things cheaper than I can? How much did you spend the materials for your gun, if you care to remember and break it down? If anyone knows where I can get anything for less, do let me know, I would appreciate it very much.
I think your buss bars could be a little wider. also, for the caps, you just linked to the BCD parts Ebay store, not the actual caps.
the caps I used (2400uf, 450v, 525v surge) have held up fine discharging short circuit, through inductive loads (with a massive clamping diode). I can't say that pulse caps wouldn't be nice, but the price per joule is so much higher. this guy is using similar 'lytics with fair success. a 24kj bank, in fact.
mounting the sparkplug will be a PITA. I was going to tap a 3/8" NPT hole in a bushing, but I decided the wall would be too thin, so I filed off the threads on the sparkplug, the dremeled threads on the bushing, and JB welded it together. so far, it has held to 5x 3 times.
do you have parts for the fuel meter? Enclosure for the caps (I put mine in a box from a case of wine)? Materials for the projectiles?
I used mcmaster mainly, but got different capacitors (BCD ones are a better deal). Why do you need so much fiberglass and teflon?
Who is online
Registered users: Bing [Bot], Majestic-12 [Bot], Yahoo [Bot]